Related Experiment Video
Updated: Jun 6, 2025
08:47
Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation
Published on: February 9, 2024
1.3K
Semantic segmentation network for mangrove tree species based on UAV remote sensing images
Xin Wang1,2,3, Yu Zhang4, Jingye Ca1
1School of Information and Software Engineering, University of Electronic Science and Technology of China, Chengdu, 610000, China.
Scientific Reports
|December 2, 2024
Summary
This study introduces UrmsNet, a new AI model for accurate mangrove species identification using RGB images, significantly reducing costs and complexity for ecological monitoring. It offers a more efficient and precise alternative to traditional remote sensing methods.
Area of Science:
- Ecology
- Remote Sensing
- Computer Vision
Background:
- Mangroves are vital coastal ecosystems requiring accurate species identification for conservation.
- Existing monitoring methods using multispectral/hyperspectral sensors are costly and complex.
- Challenges in mangrove identification include similar appearances, environmental variability, and data scarcity.
Purpose of the Study:
- To develop a cost-effective and accurate method for routine mangrove species identification.
- To improve mangrove monitoring efficiency and accessibility for ecological analysis.
- To address data scarcity in mangrove species segmentation using RGB imagery.
Main Methods:
- Proposed UrmsNet segmentation network incorporating SCConv, Adaptive Selective Attention Module (ASAM), and Cross-Layer Feature Fusion Module (CLFFM).
- Developed a high-quality RGB image dataset for mangrove species segmentation.
- Evaluated UrmsNet performance using mIoU and mPA metrics.
Main Results:
- UrmsNet achieved high accuracy with mIoU of 92.21% and mPA of 95.98% on RGB data.
- The network demonstrated comparable performance to advanced multispectral/hyperspectral methods with lower cost and complexity.
- UrmsNet showed strong performance on external datasets (LoveDA, Potsdam, Vaihingen), indicating broader applicability.
Conclusions:
- UrmsNet offers a precise, efficient, and low-cost solution for routine mangrove species monitoring.
- Combining periodic hyperspectral monitoring with UrmsNet enables comprehensive ecological assessment.
- The research provides a valuable technical approach for large-scale monitoring of coastal ecosystems.
